Glucagon-like peptide-1 receptor agonists and risk of diabetic retinopathy in type 2 diabetes: A systematic review and network meta-analysis of randomized clinical trials.

Glucagon-like peptide-1 receptor agonists (GLP-1RAs) improve cardiometabolic outcomes in type 2 diabetes, but retinal safety remains debated as use expands. We searched Embase, PubMed, Web of Science, and ClinicalTrials.gov through November 15, 2025, for randomized trials reporting diabetic retinopathy (DR) in type 2 diabetes. Bayesian random-effects network meta-analysis estimated odds ratios with 95% credible intervals. Risk of bias and certainty, effect modification, dose-response, and information sufficiency were assessed with RoB 2/CINeMA, network meta-regression, model-based network meta-analysis, and trial sequential analysis. Eighty-seven trials (9 agents; 125,418 participants) reported 6,330 DR events. No GLP-1RA was associated with a statistically credible change in DR risk versus placebo, and rankings showed no credible between-drug differences. Exploratory trial-level analyses suggested effect modification by baseline DR risk and body mass index, but not by hemoglobin A1c (HbA1c) or HbA1c change. Dose-response relationships were observed for HbA1c, weight, and blood pressure, but not for DR. Trial sequential analysis crossed the futility boundary, excluding a class-wide ±20% effect on DR risk. Current randomized trial data do not demonstrate a class-wide increase in reported DR events, although smaller, drug-specific, or high-risk subgroup effects cannot be excluded. Ophthalmic surveillance remains advisable. Trials with prespecified ocular endpoints are needed.